Every European error tracking tool reviewed
Amsterdam, Netherlands
Founded 2012
Free plan (50K requests) / €219 per year for 250K requests a month, logging add-on €89/month
30-day free trial, no credit card, plus a permanent free plan
Best for: Teams running more than one language who want errors, performance and logs in one EU-hosted tool
AppSignal is the only tool in this category that does not make you choose a runtime first.
First-party integrations cover Ruby with Rails, Sinatra and Sidekiq, Elixir with Phoenix, Ecto and Oban, Node.js with Express, NestJS and Next.js, Python with Django, FastAPI, Flask and Celery, and front-end JavaScript for React, Vue, Angular and Ember, with Go, Java, PHP and Rust arriving through OpenTelemetry. Around the error tracking sit performance monitoring, host metrics, uptime checks, log management, anomaly detection and cron monitoring, so a small team can stop at one contract.
AppSignal B.V. is registered with the Dutch Chamber of Commerce in Amsterdam under number 62960407, and the hosting is the part that separates it from most European branding: the service runs on Worldstream B.V., a Dutch hosting company, with Amazon Web Services EMEA SARL used for backups and Mailgun for email — three subprocessors in total, published.
The company states ISO 27001 and SOC 2 certification, GDPR compliance and HIPAA as a paid add-on, runs regular pentests and operates a bug bounty. An optional US region exists for customers who want one.
Pricing is per request rather than per seat, which suits teams more than it suits high-traffic single-purpose apps: the free plan covers 50,000 requests a month with 1GB of logging and five-day retention, and the paid plan is €219 a year for 250,000 requests a month, with long-term log storage a separate €89 add-on.
Users, teams, applications and dashboards are unlimited on both. The stated upgrade policy only charges after you exceed limits in two of three months, and monitoring is not cut off when you go over.
What AppSignal does well
- First-party support for Ruby, Elixir, Node.js, Python and front-end JavaScript
- ISO 27001 and SOC 2, GDPR, HIPAA available as an add-on
- Hosted on Worldstream in the Netherlands, with only three published subprocessors
- Errors, performance, logs, uptime and cron monitoring in one product
- Unlimited users and applications on every plan, including the free one
Where AppSignal falls short
- Backups sit on Amazon Web Services EMEA SARL, so the stack is not European end to end
- No self-hosting option at any price
- Per-request pricing punishes high-traffic, low-value endpoints
- Long-term log retention is a separate €89 per month add-on
- Go, Java, PHP and Rust arrive only through OpenTelemetry, not a first-party SDK
Standout feature. Three subprocessors, all named: AppSignal is the only tool here whose primary hosting is a European company rather than a European region of a US cloud.
Bonn, Germany
Founded 2014
Team from €69/month, Business from €159/month, both rising in tiers with request volume; Enterprise on request
14-day free trial, no credit card or phone number
Best for: PHP teams that need to know why a request threw and why it was slow, on a named German host
Tideways is the most specialised tool in this category and better for it.
It is built for PHP alone, extending the engine rather than sitting on top of it, which is why the profiler produces callgraphs and flamegraphs with data on SQL queries, external HTTP calls, caches and framework internals out of the box. Error and exception tracking runs alongside, so the same tool that tells you a checkout endpoint is throwing also shows the call path that got there. Laravel, Symfony, Magento, WordPress and Drupal are all covered.
Tideways GmbH sits at Königswinterer Straße 116 in Bonn, registered at Amtsgericht Bonn under HRB 22127, and has been at this since 2014. The hosting answer is the most specific in this category: everything — account data, monitoring data and profiling traces — is stored in SysEleven data centres in Berlin, Germany. No US cloud sits underneath, and the vendor names the host rather than a continent.
Pricing scales with request volume in tiers, starting at €69 a month for Team and €159 for Business at the five-million-request level, rising through nine steps to a billion requests, with 10% off for annual payment.
Team gives 5-day lookback and a single team; Business adds 30-day lookback, 15 months of history, distributed profiling, advanced alerting with PagerDuty and Opsgenie, and up to five teams. Projects, services and users are unlimited on both. There is a 14-day trial and no free plan.
What Tideways does well
- Error tracking and a real PHP profiler in the same tool
- All data stored at SysEleven in Berlin, the most specific residency claim here
- Unlimited projects, services and users on every plan
- Deep framework instrumentation for Laravel, Symfony, Magento and Drupal
- 14-day trial with no credit card or phone number
Where Tideways falls short
- PHP only — useless for any other runtime in your stack
- No free plan, and €69 a month is a high floor for a small site
- Team plan keeps only five days of lookback
- No published security certification
- Alerting on the Team plan is limited to ten transactions and a ten-minute resolution
Standout feature. A named data centre: Tideways is one of only two vendors here that says which building your stack traces are in, and the only one where nothing is subcontracted to a US cloud.
Utrecht, Netherlands
Free self-hosted with unlimited users / hosted free up to 15K events a month, then €16, €50 and €158 per month
Free self-hosted edition and a free hosted tier of 15K events a month
Best for: Teams that need error tracking on their own hardware without running a Sentry cluster
Bugsink exists because self-hosting Sentry became a project in itself. It is a single application with a database, installed by Docker, that speaks the Sentry SDK protocol — so whatever Sentry client your language already has keeps working, and the only change is the DSN. You get stack traces with local variable inspection, issue grouping and the things an error tracker is for, and none of the queueing, search and streaming infrastructure that makes the alternative a cluster.
Bugsink B.V. is at Peter Schathof 41 in Utrecht, the Netherlands, the terms are governed by Dutch law with disputes in Dutch courts, and the privacy policy states that hosted application data is stored and processed within the European Union.
The self-hosted edition is free with unlimited users and all features; the hosted plans run free to 15,000 events a month, then €16 for 75,000, €50 for 600,000 and €158 for 3 million, with premium support at €15 per user per month on top.
The licence deserves a careful read. Bugsink is source-available under the PolyForm Shield License 1.0.0, not an OSI open-source licence: you may read, modify and run it internally at no cost, but you may not use it to build something that competes with Bugsink. Parts of the repository carry other licences, including BSD-3-Clause code from Sentry itself. For internal use none of this bites; for anyone planning to offer it as a service, it is the whole document.
What Bugsink does well
- Drop-in for existing Sentry SDKs — change the DSN and nothing else
- Free self-hosted with unlimited users and no feature gate
- Deliberately simple to run: one application, one database, Docker
- Hosted tier is free to 15,000 events a month and starts at €16
- Dutch company, Dutch law, EU processing stated in the privacy policy
Where Bugsink falls short
- Source-available under PolyForm Shield 1.0.0, not open source — competing use is forbidden
- Error tracking only: no APM, no profiler, no log management around it
- The hosted service names no provider or region beyond "the European Union"
- Self-hosting means sizing for the worst deploy, not the average day
- A young single-vendor product with a small ecosystem around it
Standout feature. Sentry protocol compatibility: Bugsink is the only tool here that an already-instrumented application can adopt by changing one configuration string.
Antwerp, Belgium
Hobby €9/month, Standard €29, Growth €69, Premium €279; annually €99, €319, €759 and €3,069
10-day free trial, no credit card; no free plan
Best for: Laravel and PHP teams who want error tracking from people who write Laravel packages
Flare is built by Spatie, the Antwerp shop behind a long shelf of Laravel packages, and it shows in what the product understands.
Exceptions are handled with first-class knowledge of queues, jobs and Livewire rather than as generic stack traces, which is the difference between an error report that names the failing job and one that names the worker process. Performance monitoring for slow requests and backed-up queues and real-time log streaming sit alongside, and the front-end side covers JavaScript, React, Vue and Svelte. GitHub, GitLab, Jira and Linear are the integrations.
The operating entity is Facade BV, Kruikstraat 22/12, 2018 Antwerp, Belgium, registered under BE0729.961.919, and the terms of use state plainly that all data is processed within the European Union. What the terms do not say is which provider or which region, and the DPA is a downloadable PDF rather than a published subprocessor list — so the commitment is contractual and the detail is not public.
Pricing is a straight ladder: Hobby at €9 a month for one user and one project, Standard at €29 with unlimited users and projects, Growth at €69 and Premium at €279 for high-volume work, with annual payment at €99, €319, €759 and €3,069 respectively. The trial is 10 days with no credit card, and there is no free plan — the shortest trial and the lowest paid entry point in this category at the same time.
What Flare does well
- Laravel-native: understands queues, jobs and Livewire rather than generic traces
- Errors, performance monitoring and log streaming in one product
- Front-end coverage for JavaScript, React, Vue and Svelte
- EU-only processing written into the terms of use
- Entry plan at €9 a month, the lowest paid floor here
Where Flare falls short
- No free plan and only a 10-day trial, the shortest in this category
- No provider or region named behind the EU commitment
- Hobby plan is capped at one user and one project
- PHP and front-end JavaScript only — nothing for Ruby, Python, .NET or Go
- No published security certification
Standout feature. Built by Spatie: the Laravel handling comes from the team whose packages are already in the application you are instrumenting.
Laasby, Denmark
Small Business $26/month, Business $44, Business+ $89, Enterprise $269, always billed in USD
21-day free trial, no credit card; no free plan
Best for: .NET teams that want error logging, uptime and heartbeats from one Danish vendor
elmah.io is the commercial descendant of ELMAH, the error logging library that a generation of ASP.NET applications shipped with, and it has stayed in that lane.
It supports every .NET web and logging framework, including ASP.NET Core and .NET Lambda, plus JavaScript, and adds the pieces a .NET shop otherwise buys separately: uptime monitoring with SSL certificate validation, heartbeat monitoring for scheduled jobs, deployment tracking across versions, and bot detection to keep crawler noise out of the issue list. An AI assistant explains stack traces and suggests fixes, and an MCP server exposes it to editors like Cursor.
elmah.io ApS is at J M Knudsens Vej 14, 8670 Laasby, Denmark, and the terms are governed by Danish law. The security page states PCI DSS compliance, GDPR, DDoS protection, end-to-end 256-bit TLS, continuous penetration testing and two-factor authentication.
It also states that the service runs on Microsoft Azure — and stops there. Which Azure region holds your exceptions is not published anywhere on the site, which is a gap worth closing with the vendor before signing if residency is in your requirements.
Pricing runs Small Business at $26 a month, Business at $44, Business+ at $89 and Enterprise at $269, with a 21-day trial, no credit card and a 14-day money-back guarantee. Note the currency: a Danish company that always bills in USD, with tax added in USD for some countries, which puts an exchange-rate line into a European budget for no obvious reason. Discounted or free access exists for startups, non-profits, user groups and open-source projects.
What elmah.io does well
- Covers every .NET web and logging framework, plus JavaScript
- Uptime, SSL and heartbeat monitoring bundled with error logging
- Danish company under Danish law, PCI DSS and GDPR stated
- Bot detection keeps crawler noise out of the issue list
- 21-day trial, no card, plus free access for open source and non-profits
Where elmah.io falls short
- Names Microsoft Azure as its platform but never states the region
- .NET and JavaScript only
- Always billed in USD despite being a Danish company
- No free plan
- Azure underneath means US cloud ownership regardless of the Danish letterhead
Standout feature. The ELMAH lineage: for a .NET team, this is the hosted version of the library the codebase probably already used.
Naples and Florence, Italy
Free up to 30K transactions a month / Beginner €15, Developer €39, Team €69 per month
Free plan with no time limit and no card required
Best for: Small PHP teams that want real monitoring on a free tier that never expires
Inspector is the cheapest honest way into this category. The free plan is 30,000 transactions a month with unlimited events, hosts and applications and up to three team members, and the vendor is explicit that it is a monthly quota rather than a countdown — you can run it on a production workload indefinitely without paying.
Above that, Beginner is €15 a month for 300,000 transactions, Developer €39 for 800,000 and Team €69 for 1.5 million, all with unlimited users, REST API access and live chat support.
The product is real-time monitoring for PHP applications with errors and performance together, covering Laravel, Symfony, CodeIgniter and Drupal, plus the Neuron AI framework, with an AI bug fixer on the paid tiers and zero-configuration installation. Inspector S.R.L. is an Italian company with a VAT number of 09552901218 and offices at Via Guantai ad Orsolone 178 in Naples and Via Ippolito Pindemonte 6 in Florence, started with support from the Italian government's Resto al SUD scheme.
The gap is documentation of the things this directory exists to check. Inspector publishes no data residency statement, names no hosting provider or region, and lists no security certification on the pages checked. The company is demonstrably Italian, so the jurisdiction question is settled; where the servers are is not, and for a regulated buyer that is a question to put in writing before signing rather than an assumption to make from the address.
What Inspector does well
- Free plan of 30,000 transactions a month with no time limit
- Unlimited users on every paid plan, from €15 a month
- Laravel, Symfony, CodeIgniter and Drupal covered out of the box
- Errors and performance monitoring in one product
- Italian company with published offices in Naples and Florence
Where Inspector falls short
- No data residency statement anywhere on the site — no provider, no region
- No published security certification
- PHP-centric; other runtimes are not covered
- Free plan caps the team at three members
- Smallest vendor here, with the thinnest public documentation
Standout feature. A free tier that is a quota, not a trial: Inspector is the only paid-tier vendor here that lets a production application run on the free plan forever.
Walldorf, Germany
Free plan (100K log lines a day, 2,000 devices) / Team from $39/month, Pro from $89/month; overage €3.25 per extra million logs
Free plan, no credit card
Best for: Mobile teams that need the log leading up to the crash, not just the crash
Bugfender inverts the usual model. Instead of waking up when an exception is thrown, it logs continuously — 24 hours a day, whether or not anything has gone wrong — and keeps those logs per device, so when a user reports that something behaved strangely you can read what the app actually did rather than guess.
Crash reporting sits on top, with stack traces and automatic symbolication, alongside a user feedback channel that arrives with the device context already attached. That combination is aimed squarely at the mobile bug that does not crash.
Coverage is iOS, Android, JavaScript, React, Angular, Flutter, Unity, Ionic, Xamarin, Cordova and PhoneGap. The operator is Beenario GmbH, Altrottstraße 31, 69190 Walldorf, Germany, VAT DE299463958, registered at Amtsgericht Stuttgart under HRB 752438. The privacy policy is unusually specific about subprocessors: data storage is provided by Wasabi Technologies, Inc. with the data stored in the EU, with StatusPage.io for incidents, Cloudflare for traffic filtering and Intercom for chat. ISO 27001 and GDPR compliance are stated.
Pricing is by log volume, which is the meter to model before signing.
The free plan covers 100,000 log lines a day across up to 2,000 devices with 24-hour retention; Team starts at $39 a month for 10 million monthly log lines with 7-day retention and unlimited devices; Pro starts at $89 with 30-day retention; Enterprise is custom and is where on-premises deployment and a choice of server geography live. Overage runs €3.25 per additional million logs, billed in one-million batches, and annual billing takes 30% off.
What Bugfender does well
- Continuous remote logging, not only crash capture
- Crash stack traces with automatic symbolication
- Eleven platforms covered, including Flutter, Unity and Xamarin
- German company, ISO 27001 stated, EU data storage
- Free plan with 100,000 log lines a day and unlimited app installs
Where Bugfender falls short
- EU storage sits on Wasabi Technologies, Inc., a US corporation
- Free plan retains logs for only 24 hours
- Priced per log line, so a chatty debug build is a budget event
- Choosing your server geography is an Enterprise-only option
- Mobile and front-end oriented; not a back-end error tracker
Standout feature. The log before the crash: Bugfender is the only tool here that records what a device was doing when nothing had gone wrong yet.
London, United Kingdom
Flex private cloud from $175/month, usage-based; Enterprise self-hosted on request
14-day free trial of Flex
Best for: Mobile product teams that want crash data and usage analytics in the same system, self-hosted if needed
Countly treats crashes as one module inside a product analytics platform rather than as the product. That is either the point or the problem depending on who is buying: a product manager gets crash rates sitting next to funnels, retention, cohorts and A/B tests, so "this release crashes" and "this release converts worse" are the same chart. An engineer looking only for stack traces is paying for a great deal of machinery they will not open.
Countly Ltd is incorporated in England and Wales under number 8568329, with a registered office at 1 Bow Churchyard, London EC4M 9DQ and VAT number GB168599344.
Ten SDKs cover iOS, Android, Flutter, JavaScript, React, Node.js, Windows, C++, Java and Unity, with a write API for anything else. Hosting is Google Cloud, in a region the customer picks, with EU regions available; the Enterprise edition is deployed on customer infrastructure with the data location fully customer-defined and, as the vendor puts it, zero vendor access.
That self-hosted route is the reason Countly appears in this directory at all: it is one of only two tools here that can run entirely on your own machines. The cost of entry is high.
Flex private cloud starts at $175 a month on usage-based pricing with a 14-day trial, and Enterprise is quoted rather than listed. Subprocessors for the hosted service include OpenAI for the AI assistant, Intercom for support, Postmark for email and Paddle for payments — a longer list than the specialists here publish.
What Countly does well
- Crash and error data alongside funnels, retention and cohorts
- Ten SDKs across mobile, web and desktop, plus a write API
- Enterprise edition self-hosted with customer-defined data location
- Choose your Google Cloud region, including EU regions
- UK company with a published registered office and VAT number
Where Countly falls short
- Hosted service runs on Google Cloud, a US provider
- Starts at $175 a month, by far the highest floor in this category
- An analytics platform first; error tracking is one module of many
- Longer subprocessor list than the specialists, including OpenAI and Intercom
- UK jurisdiction rather than EU, so the transfer question is adequacy rather than intra-EEA
Standout feature. Crashes inside the analytics: the only tool here where a crash rate and a conversion rate can be plotted against the same release.
Schwarzach, Austria
Starter $49/month billed annually ($59 monthly), Team $149, Pro $299, Enterprise from $999
14-day free trial; no free plan
Best for: Product teams that want user-reported bugs to arrive with the console log already attached
Gleap is not an error tracker and should not be bought as one.
It is an in-app feedback and bug reporting layer: a user presses a button, and the report arrives with console logs, a screenshot, session replay and the technical context around the moment, so the ticket is reproducible instead of being three sentences and a guess. That closes a real gap — the bug a user can see and the tracker cannot, because nothing threw — but it only ever fires when a human decides to press the button.
Gleap GmbH is at Am Dorfplatz 3, 6858 Schwarzach, Austria, registered at the Regional Court of Feldkirch under FN 534390 v with VAT ATU 75671417. The hosting is described precisely in the privacy policy: the main infrastructure and production MongoDB database run on DigitalOcean managed databases and application servers in Frankfurt, Germany, with Cloudflare providing static file hosting for screenshots and attachments and the video-call infrastructure. Azure and AWS are named as providers no longer used.
The AI features are where the subprocessor list gets long and worth reading before enabling them: ElevenLabs, MongoDB/Voyage AI, OpenAI Ireland Ltd and Grok are routed through EU endpoints, but the policy states that Anthropic may process in the US or the EU without an EU-only guarantee, OpenRouter may route to the EU, US or China, and DeepSeek processes in China.
Pricing starts at $49 a month billed annually for Starter, $149 for Team, $299 for Pro and from $999 for Enterprise, with a 14-day trial and no free plan.
What Gleap does well
- Bug reports arrive with console logs, screenshot and session replay attached
- Servers in Frankfurt, Germany, named down to the provider and database
- Austrian GmbH with full register and VAT details published
- Covers websites and mobile apps from one SDK
- Unusually candid privacy policy about where each AI subprocessor runs
Where Gleap falls short
- Not an error tracker: nothing is captured unless a user reports it
- Infrastructure is DigitalOcean and Cloudflare, both US companies
- Some AI features route to Anthropic, OpenRouter or DeepSeek outside the EU
- No free plan and $49 a month is a high floor for the job it does
- No published security certification
Standout feature. A precise hosting statement: Gleap names the provider, the service and the city, which most vendors in this category do not.
Brussels, Belgium
Starter $39/month billed annually ($59 monthly), Team $149 ($199 monthly), Business on request
15-day free trial, no credit card; no free plan
Best for: Agencies and QA teams pushing client-reported website bugs into an existing issue tracker
Marker.io sits between the person who found the bug and the developer who has to fix it.
A reporter annotates the page, and the issue lands in Jira, Linear, GitHub or the tracker the team already uses, carrying session replay, console logs and the technical context of the session rather than a screenshot pasted into an email. For agencies collecting feedback from clients who will never have a tracker login, that routing is the entire product, and it is done well.
Marker.io SRL is a Belgian company registered under number 0556685968 with its registered office at Avenue Louise 231, 1050 Brussels. The DPA states that hosting is on AWS and describes it as ISO 27001-certified infrastructure — which is AWS's certificate rather than Marker.io's, and no region is named. So the company is Belgian and the servers are a US cloud in an unstated location: the ownership question is answered and the residency question is not.
Pricing is $39 a month for Starter billed annually ($59 monthly) with three users and unlimited feedback, $149 for Team ($199 monthly) with fifteen users, and Business on request with yearly billing only. An agency plan gives fifteen members, fifty projects and fifty guests at $99 a month annually. The trial is fifteen days with no credit card, all prices are in USD, and there is no free plan.
What Marker.io does well
- Reports land directly in Jira, Linear, GitHub or your existing tracker
- Session replay and console logs attached to each report
- Guests can report without a licence in your issue tracker
- Belgian company with published registration and registered office
- 15-day trial with no credit card, plus a dedicated agency plan
Where Marker.io falls short
- Not an error tracker: nothing is captured unless a human submits it
- Hosted on AWS with no region stated anywhere
- The ISO 27001 claim belongs to AWS, not to Marker.io
- Priced in USD with no free plan, from $39 a month for three users
- Web only — no mobile SDK
Standout feature. It never becomes a second backlog: Marker.io deliberately files into the tracker you already run rather than asking the team to watch another inbox.
